The objective of the proposed research is to test two hypotheses for the failure of BCG Vaccination to protect against tuberculosis in a rural population in South India, an area where tuberculosis is caused by tubercle bacilli of low virulence for humans. These strains of tubercle bacilli are also of low virulence for guinea pigs. The hypotheses to be tested relate to the influence of infection with atypical mycobacteria on protection induced by BCG vaccine. Guinea pigs infected with Mycobacterium avium-intracellulare will be injected with BCG or placebo. The animals will later be infected by the respiratory route with 5-10 tubercle bacilli of strains differing in virulence. At intervals thereafter, animals will be killed in order to quantitate the number of tubercle bacilli recovered from the lungs and spleen.